SKILL: Customer Spend Monitor

Name: Skill
Rating: 92
Author: team-telnyx

Name

customer-spend-monitor

Description

Daily customer revenue pace monitor. Compares current MTD revenue against prorated prior month for each customer, flags significant changes, and surfaces a watch list of customers with escalations or stale tickets. Posts alerts to Slack and/or Telegram.

Schedule

Cron: 43 8 * * 2,4 (8:43 AM CT on Tuesdays and Thursdays)

Commands

bash

# Full run — posts to all configured channels
bash scripts/customer-spend-monitor.sh

# Dry run — calculate and print, don't post
bash scripts/customer-spend-monitor.sh --dry-run

# Save output to file
bash scripts/customer-spend-monitor.sh --output report.txt

# Single customer only
bash scripts/customer-spend-monitor.sh --customer "Ringba"

# Post to Slack only
bash scripts/customer-spend-monitor.sh --slack-only

# Post to Telegram only
bash scripts/customer-spend-monitor.sh --telegram-only

Environment Variables

Variable	Required	Description
`TABLEAU_PAT_SECRET`	Yes	Tableau Personal Access Token secret
`SLACK_BOT_TOKEN`	Yes*	Slack bot token with `chat:write` scope
`TELEGRAM_BOT_TOKEN`	No	Telegram bot token (if using Telegram)
`CONFIG_PATH`	No	Path to config JSON (default: `./config/config.json`)
`DRY_RUN`	No	Set `true` to skip posting

* Required unless using --telegram-only

Built-in Resilience

•Config validation at startup — catches missing/invalid settings early
•Retry logic — 3 attempts with exponential backoff on transient failures
•Curl timeouts — 10s connect, 30s max per request
•Tableau fallback — falls back to billing A2A agent if Tableau is unreachable
•Dry-run mode — preview output without posting

Dependencies

•bash, curl, jq, bc
•Network access to Tableau Server
•Network access to revenue-agents.query.prod.telnyx.io:8000 (fallback)

Author

team-telnyx / CSM team